You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@jena.apache.org by Adrian Gschwend <ml...@netlabs.org> on 2025/02/03 14:50:48 UTC
Re: Feature Request: Filtered Output Option for riot --validate
On 26.01.2025 19:13, Andy Seaborne wrote:
Hi Andy,
> Aside: do you use SHACL?
I do but for other use-cases. In this particular use-case, I was
exporting a large government LOD database and we don't have influence on
all the stuff some pipelines write into it. And I simply wanted to run
it on another database.
> Recovery means reworking parsers. Which RDF syntaxes are you interested
> in? N-triples can make use of the statement per line. Turtle can't but
> recovery might risk skip-to-DOT. RDF/XML is based on an XML parser.
> JSON-LD is a 3rd party engine.
ah good point, I did not think that through. I personally have N-Triples
in this scenario but I completely see your point, in another, similar
use-case the source was a ton of JSON-LD files that were sometimes
simply broken.
That brings me to another issue I had with this JSON-LD dump: It was
around 60k files so shell expansion of *.json failed on all shells I
tried due to the size and I could not find a glob option in riot. And
scripting it didn't work either due to JVM startup time.
> Warnings - do you mean IRI warnings? - or some syntax level warnings?
IRI warnings or invalid datatypes, usually some xsd that is wrong. 31st
of February was a common one in this dataset for reasons.
> There's a new IRI subsystem in the pipeline - and its errors/warnigns
> are more controllable -
> > https://lists.apache.org/list.html?
> users@jena.apache.org#:~:text=Fuseki%20development%20features.-,
> %3D%3D%3D%3D%20IRI3986,-Issue%3A%20https
was that a direct link to a post on the list? I get only an index.
regards
Adrian
Re: Feature Request: Filtered Output Option for riot --validate
Posted by Adrian Gschwend <ml...@netlabs.org>.
On 05.02.2025 02:54, Andy Seaborne wrote:
> Someone is looking at a native of Jena which would be great for command
> line use.
>
> https://github.com/apache/jena/discussions/2934
that's cool, would definitely be helpful for everything cli.
regards
Adrian
Re: Feature Request: Filtered Output Option for riot --validate
Posted by Andy Seaborne <an...@apache.org>.
On 03/02/2025 14:50, Adrian Gschwend wrote:
> On 26.01.2025 19:13, Andy Seaborne wrote:
>
> Hi Andy,
>
>> Aside: do you use SHACL?
>
> I do but for other use-cases. In this particular use-case, I was
> exporting a large government LOD database and we don't have influence on
> all the stuff some pipelines write into it. And I simply wanted to run
> it on another database.
>
>> Recovery means reworking parsers. Which RDF syntaxes are you
>> interested in? N-triples can make use of the statement per line.
>> Turtle can't but recovery might risk skip-to-DOT. RDF/XML is based on
>> an XML parser. JSON-LD is a 3rd party engine.
>
> ah good point, I did not think that through. I personally have N-Triples
> in this scenario but I completely see your point, in another, similar
> use-case the source was a ton of JSON-LD files that were sometimes
> simply broken.
>
> That brings me to another issue I had with this JSON-LD dump: It was
> around 60k files so shell expansion of *.json failed on all shells I
> tried due to the size and I could not find a glob option in riot. And
> scripting it didn't work either due to JVM startup time.
Someone is looking at a native of Jena which would be great for command
line use.
https://github.com/apache/jena/discussions/2934
>
>> Warnings - do you mean IRI warnings? - or some syntax level warnings?
>
> IRI warnings or invalid datatypes, usually some xsd that is wrong. 31st
> of February was a common one in this dataset for reasons.
>
>> There's a new IRI subsystem in the pipeline - and its errors/warnigns
>> are more controllable -
> > > https://lists.apache.org/list.html?
>> users@jena.apache.org#:~:text=Fuseki%20development%20features.-,
>> %3D%3D%3D%3D%20IRI3986,-Issue%3A%20https
>
> was that a direct link to a post on the list? I get only an index.
>
> regards
>
> Adrian
Re: Feature Request: Filtered Output Option for riot --validate
Posted by Andy Seaborne <an...@apache.org>.
On 03/02/2025 14:50, Adrian Gschwend wrote:
> That brings me to another issue I had with this JSON-LD dump: It was
> around 60k files so shell expansion of *.json failed on all shells I
> tried due to the size and I could not find a glob option in riot. And
> scripting it didn't work either due to JVM startup time.
xargs(1) chops up inline into lengths that will fit the command line limits:
find . -name '*.json' | xargs riot
>> Warnings - do you mean IRI warnings? - or some syntax level warnings?
>
> IRI warnings or invalid datatypes, usually some xsd that is wrong. 31st
> of February was a common one in this dataset for reasons.
As is "tuesday"^^xsd:date
>
>> There's a new IRI subsystem in the pipeline - and its errors/warnigns
>> are more controllable -
> > > https://lists.apache.org/list.html?
>> users@jena.apache.org#:~:text=Fuseki%20development%20features.-,
>> %3D%3D%3D%3D%20IRI3986,-Issue%3A%20https
>
> was that a direct link to a post on the list? I get only an index.
Try this:
https://lists.apache.org/thread/4gk0kpf75hyk58db343hx13nzq8pvldg
and go down to "==== IRI3986"
Andy